Modelagem do processo de autoprecipitação de gibsita em decantadores do processo Bayer

Abstract

Most Bayer process refineries monitor carefully alumina losses. One of the prominent effects that cause it is auto precipitation in the clarification circuit. This document proposes a conceptual model for the representation of the process nuances, considering settler solids profile and composition, and the uneven fluid flow behavior of a vessel of this type. The technique consists in discretization of of bayer process decanters in stagest with representation of mixing, reaction, and splitting. It is proven to be at zero degrees of freedom independently of the composition detail on liquor and solids composition through a generalized degrees of freedom analysis for an undefined number of components and stages. The effect of parameters such as number of stages, feed stream position, and mud bed level on the model response and stability is analyzed so to propose an ideal range of numerical modelling conditions. The proposed model was shown to be robust and has guaranteed a steady state solution attaining results that are in accordance with reality. Additionally, the impact of the number of discretization does not add to the model capacity, in fact it is detrimental to it by adding numerical instability while the combination of number of stages and position of the feed possess an optimal functioning combination.